This study proposes to conduct a meta-analysis of a substantial library of reports of juvenile delinquency treatment effectiveness research. Meta-analysis is a statistical technique for quantitatively synthesizing and analyzing the results of treatment studies. It is based on a detailed coding of the characteristics of each eligible study and the treatment outcome effect size on each dependent variable of interest in the study. The goals of this meta-analysis are to: (1) Estimate the mean overall treatment effect size and its statistical confidence limits for the population of relevant and methodologically credible studies of the effects of delinquency treatment and examine differential treatment effects on different classes of outcome variables: recidivism, delinquent behavior, attitudes, achievement, self-esteem, and so forth. (2) Analyze the relationship between the effects of delinquency treatment and such circumstances of treatment as type of client, treatment modality, treatment intensity, setting, and so forth in order to identify any factors associated with treatment efficacy. (3) Analyze the relationship between the methodology used in studies of delinquency treatment and the resulting effect sizes in order to identify those methodological practices that introduce bias or excessive variability into the results. Eligible research reports for meta-analysis are defined as those that study the effects of treatment on juveniles, involve delinquent subjects or delinquent behavior as an outcome variable, and use a quantitative comparison group methodology (randomized experiment or quasi-experiment). A bibliography of over 7000 studies of delinquency treatment has been compiled from computer and manual searches. Of these, 2163 reports have been identified from their abstracts or the full report as possible candidates for paper copy or microfiche form. Screening and retrieval are underway for additional candidate reports. It is estimated that, after close inspection for eligibility and completeness, 800-1000 reports will be codable for meta-analysis. The present proposal seeks funding to complete the retrieval of reports, perform the necessary coding, analyze the resulting quantitative data, and report the results.